Abstract (EN):
In this paper, automated guided vehicles (AGVs) control in indoors quasi-structured environments is considered where mismatched wheels radius systematical errors corresponding to differential-drive based AGVs are presented. The main wheels radius mismatch effect stems from wearage and mechanical imperfections and tolerance, and it leads to deadreckoning navigation performance decrement. Each AGV follows trajectories composed of lines and changes of direction events, because their path planning can be done off-line due to the quasi-structured environment condition. Past work was published where systematical errors had AGV dead-reckoning estimated pose influence on system accuracy. In this paper, these results are further demonstrated through an experimental setup in which an AGV is controlled to follow a trajectory and dead-reckoning induced systematical errors are computed. The results confirm that heading measurement accuracy can greatly influence dead-reckoning performance even in the presence of strong detrimental AGV errors effects.
Language:
English
Type (Professor's evaluation):
Scientific
Contact:
azenha@fe.up.pt